1. ACEA PM-3 programme Caroline Hosier, Pilot, ACEA Task-Force Particulates (TF-PM) Ford Motor Company Draft presentation to GRPE Informal document No. GRPE-52-9 (52nd GRPE, 6-9 June 2006, agenda item 2.)

2. 2 Background The ACEA 1st and 2nd Particulate programmes concentrated on particulate measurement of mass, size, number and characterised particulate emissions from diesel, gasoline, diesel with trap and direct injection petrol engines. The ACEA 3rd particulate programme, known as PM-3 was designed to contribute to the UK Government led Particulate Measurement Programme (PMP) being run under the auspices of the UN-ECE.

3. 3 Background PM-3 focuses on the gravimetric measurement method. PM-3 was initiated and funded by ACEA, using vehicles loaned from Fiat, PSA and VW. The 4th vehicle was loaned to the programme by Toyota. All testing was conducted at an independent laboratory, AVL-MTC in Sweden. The testing was completed in 2-phases. –The main programme –The recovery programme

4. 4 Objectives Primary objective: Examine potential enhancements to the existing particulate gravimetric measurement method. 1.Increase measurement filter loading. 2.Reduce Variability. Secondary objective: Investigate the potential of alternative dynamic mass- based particulate measurement methods.

5. 5 Main Test outline 5 test sets using 4 particulate probes in parallel Filters weighed on both 1.0  g and 0.1  g balances. 4 vehicles - 3 * diesel, 1 * gasoline Gasoline fuel EN228 S<10 ppm Diesel Fuel - Swedish class 1, S<10 ppm

6. 6 Main prog Probe/filter set-up Probe 1Probe 2Probe 3Probe 4 Set 1: Low dilution TX-40 + new filter holder + thermostatic control 90 l / min flow T-60 + old filter holder 30-40 l / min flow TX-40 + new filter holder without thermostatic control 90 l / min flow TX-40 + old filter holder w/o thermostatic control 30-40 litres/min Set 2: normal (high) dilution TX-40 + new filter holder + thermostatic control 90 l / min flow T-60 + old filter holder 30-40 l / min flow TX-40 + new filter holder without thermostatic control 90 l / min flow TX-40 + old filter holder w/o thermostatic control 30-40 litres/min Set 3: Testing highest then lowest emitter diesel TX-40 + new filter holder + thermostatic control 90 l / min flow T-60 + old filter holder 30-40 l / min flow TX-40 + new filter holder without thermostatic control 90 l / min flow TX-40 + old filter holder w/o thermostatic control 30-40 litres/min

7. 7 Probe/filter set-up 2 Probe 1Probe 2Probe 3Probe 4 Set 4: Mixing Tee close to tailpipe TX-40 + new filter holder + thermostatic control 90 l / min flow T-60 + old filter holder 30-40 l / min flow TX-40 + new filter holder without thermostatic control 90 l / min flow TX-40 + old filter holder w/o thermostatic control 30-40 litres/min Set 5: normal (high) dilution Teflo 2 micron + new filter holder + thermostatic control 90 l / min flow T-60 + old filter holder 30-40 l / min flow Teflo 2 micron + new filter holder without thermostatic control 90 l / min flow Teflo 2 micron + old filter holder w/o thermostatic control 30-40 litres/min

8. 8 Recovery Programme Owing to flow rate controller errors in the original programme, it was necessary for ACEA and AVL-MTC together to run a recovery test programme. This was conducted using a single PSA diesel vehicle equipped with a DPF.

9. 9 Recovery programme - 1

10. 10 Recovery programme - 2

11. 11 Probes mounted in tunnel. Tunnel End-plate with mounted Probes

12. 12 New filter holder (as specified in US HD2007) New filter holder with thermostatically controlled heating jacket

13. 13 Filter media – teflo, TX-40 & T60 Results from main programme

14. 14 Filter material (main prog.) No statistical difference in average value TX-40 to T60. No difference in Standard deviation.

15. 15 Teflo filters –Require improved handling practices as they are very susceptible to damage. –Do not easily fit the filter cassettes. –Adequate charge neutralisation is essential Teflo filter showed no difference in measurement, therefore it should be allowed as an alternative.

16. 16 Microbalance At very low PM yield, the 0.1  g balance took over 20 minutes to stabilise in some cases, particularly with the teflo filter. 1  g balance is sufficient

17. 17 Filter Holder No data available from 1 st programme on filter holder effect Data from recovery programme shows no effect of changing filter holder to the US HD-2007 design

18. 18 Filter Cassette When heavily loaded, cassette accumulates PM on the walls of the cassette itself. Difficult to retain all the PM from such a sample for weighing. The cassette has to be cleaned before being reused to relieve it from residual PM. Cassette itself causes some handling difficulties. Redesign of cassette may be required.

19. 19 Thermostatic Control Thermostatic control –of the sampling line and filter holder to 47°C± 5°C for US HD-2007 filter holder. Thermostatic control of the sampling line and the filter holder does not improve repeatibility. This was confirmed in the recovery programme

20. 20 Thermostatic Control (2) Thermostatic control of filter holder / sample lines makes no difference

21. 21 Weighing Room Experimentation on effect of weighing room environmental conditions (temperature/humidity) on blank and loaded filter papers. WAITING FOR UPDATED INFO FROM MTC

22. 22 No improvement of repeatibility with increasing sample flow Note T60 measurement filters broke at higher flows for the higher PM vehicle. Sample Flow Rates

23. 23 Conclusions 1 - draft The PM mass measurement method is suitable for measurement of very low PM emissions. Use of a high efficiency filter (TX40 or teflo) is recommended. Thermostatic control of the filter holder/sample probe makes no difference and does not improve variability. Filter holder design makes no difference and does not improve variability.

24. 24 Conclusions 2 - draft The filter cassette causes some handling difficulties and may introduce errors. Flow rate makes no different The 1.0  g balance performs better in a working environment than the 0.1  g balance.

25. 25 Conclusions 3 - draft Measurement of Particulate mass for DPF equipped vehicles requires careful handling and control. The US HD-2007 adapted method adds more complication without offering any benefits. This study did not show a statistically significant improvement when using a single TX-40 instead of the T60 (primary + secondary) filter pair. The application of a single TX-40 offers the advantage of reduced weighing activities without influencing the measured results.

26. 26 Conclusions 4 - draft A high number of tests (approximately 20%) are not valid for various reasons, even though this is a well understood method used by a reputable laboratory. However, the mass measurement method is suitable for measurement of PM from DPF equipped vehicles.

27. 27 Recommendations - 1 Modification to existing method –Single TX-40 or teflo filter (ie one filter per NEDC) Retain existing method in following respects –Existing filter holder –No heating requirement –No filter cassette –1  g balance

28. 28 Recommendations - 2 Other aspects –No benefit from increased sample flow rate –No benefit from decreased CVS flow rate –Good control of weighing room conditions essential

29. 29 Suggestions for PMP Further work Before implementing mass measurement measures from PMP, it needs to be clearly demonstrated that they give improvements. Inconsistencies which we have seen from different test programmes highlight the need for further work. In such a case, ACEA would be prepared to participate (comparison of different filter materials).